fix(workflows): reject non-string 'condition' in if/while/do-while steps

[jawwad-ali]

Problem

IfThenStep, WhileStep, and DoWhileStep validate() confirm that condition is present but never that it is a string.

At run time execute() feeds condition to evaluate_condition():

def evaluate_expression(template, context):
    if not isinstance(template, str):
        return template   # non-string returned as-is
    ...

def evaluate_condition(condition, context):
    result = evaluate_expression(condition, context)
    ...
    return bool(result)

So a non-string condition is returned unchanged and bool()-coerced. A list/dict/number authoring mistake silently resolves to a truthiness instead of erroring:

• condition: [1, 2] → bool([1, 2]) → always True → if-branch always taken / while-loop spins to max_iterations.
• condition: {} → always False, etc.

No error is reported, so the mistake is invisible.

Fix

Reject a present-but-non-string condition in each of the three steps' validate(), mirroring the existing prompt/shell/command 'must be a string' guards. "true"/"false" and expressions like "{{ inputs.flag }}" are strings, so they remain valid — no behaviour change for correct workflows.

Verification

• New parametrized tests ([list, dict, int, bool]) across all three step classes: fail before the guard (12 failures — the non-string conditions pass validation), pass after.
• test_validate_accepts_string_condition confirms "true"/"false"/"{{ ... }}" still validate.
• Full TestIfThenStep/TestWhileStep/TestDoWhileStep: 53 passed.
• ruff check clean on all changed files.

---

AI-assisted: authored with Claude Code. I traced the runtime path (evaluate_condition → bool()), confirmed the silent-coercion on main, and verified the guards fail-before/pass-after.